Every few months, a new wave of AI announcements rolls in and the discourse kicks off again: developers are done, designers are obsolete, just prompt your way to a product. I've been building software for a few years now. And honestly? I'm not scared. Here's why.
The thing people keep getting wrong
When calculators arrived, mathematicians didn't disappear — the bar for what counted as "math work" just shifted upward. The same thing is happening now. AI is eating the bottom of the stack: boilerplate, scaffolding, repetitive transformations. What's left — and what keeps growing in value — is judgment. Knowing which problem to solve. Knowing when the architecture is wrong before you've written a line. Knowing that the feature the PM asked for will create three worse problems downstream.
The question isn't whether AI can write code. It's whether AI can decide what's worth building.
— Claude ;) (the irony I know)
What actually changed for me
I write less boilerplate. I spend more time in the problem space and less time in the syntax space. My output per hour has gone up — which, if I'm being honest, just means I take on harder problems now. The fear mostly comes from people who equate coding with typing. If your mental model of software development is "person produces lines of code", then yes, that person is in trouble. But if you think of it as "person solves problems and code is one of the outputs", the picture looks very different.
So, will you lose your job?
If your only skill is producing code on demand: maybe, eventually. If your skill is understanding systems, communicating tradeoffs, and knowing what the right thing to build is: you're going to be more valuable, not less. The tools just got better. Stay curious. Keep shipping. Don't mistake anxiety for prophecy.